Background
Rabies virus (RABV) is a highly neurotropic virus that causes lethal central nervous system (CNS) disease in many species of mammals including humans [
1]. Although rabies has been well controlled in the United States and other developed countries by vaccination in animals, it is still a public health threat, causing more than 55,000 human deaths worldwide each year [
2]. Furthermore, no therapy has proved effective to cure rabid patients once rabies encephalitis develops or once the clinical symptoms appear.
Immune responses and CNS dysfunction are two main factors to be considered during RABV infection. Although RABV infection is invariably lethal in the absence of protective immune responses, several studies have argued that excessive immune responses may not always be beneficial for RABV infection. Attenuated RABV activates innate immune responses and induces extensive inflammation, apoptosis and neuronal degeneration in the CNS in experimental animals [
3]-[
6]. Moreover, the expression of the genes involved in innate immune and antiviral responses were highly upregulated after infection with attenuated RABV, especially those related to the alpha/beta interferon (IFN-α/β) signaling pathways, inflammatory cytokines and chemokines, including interleukin-6 (IL-6), IL-1α/β, IL-10, CXCL10/IP-10 and CCL5/RANTES [
7]-[
9]. However, it has been shown that overexpression of these chemokines (such as CXCL10 and CCL5) is closely correlated with severe enhancement of blood-brain barrier (BBB) permeability and excessive infiltration and accumulation of inflammatory cells in the CNS, which contributes to the increased pathogenicity in neurological diseases [
10]-[
12].
Most street RABVs evade the host innate immune system and fail to induce protective virus neutralizing antibody (VNA) responses [
13]-[
16]. However, in some murine or dog experimental models infected with street RABVs, T cell and mononuclear cell infiltration in the CNS have been observed together with severe encephalitis in the late stage of infection [
16]-[
18]. Although inflammatory response in the early stage of infection is important for clearance of RABV from the CNS [
19], there is no evidence to suggest that severe inflammation in the late stage is beneficial to or impedes the development of the disease.
Chemokines have been originally identified as chemotactic and pro-adhesive cytokines by their interaction with G protein-coupled receptors. CCL5 (also termed as RANTES) is a β chemokine and induces leukocyte migration by binding to CCR1, CCR3 or CCR5 [
20],[
21]. An elevated level of CCL5 has been associated with a variety of inflammatory disorders [
22],[
23]. As one of the CCL5 receptors, CCR5 also has a significant role in various diseases, such as AIDS [
24], arthritis [
25],
Toxoplasma gondii infection [
26], West Nile virus infection [
27] and respiratory virus infection [
28]. Met-CCL5, an N-terminally modified human CCL5, has been previously shown to inhibit activity at two rodent chemokine receptors CCR1 and CCR5 [
29]. Targeting CCL5 or CCR5 with antagonists may have potential therapeutic usage to alleviate symptoms of these diseases [
30],[
31].
In this study, mice infected with attenuated RABVs developed excessive inflammation in the CNS. CCL5 was the highest virus-induced chemokine among 40 inflammatory cytokines and chemokines, which promoted migration of macrophages and T cells into the CNS. Excitingly, administration of the CCL5 antagonist, Met-CCL5, alleviated rabies clinical symptoms and prolonged the survival time of the sucking mice infected with attenuated RABV or adult mice infected with attenuated or street RABVs. Met-CCL5 treatment significantly reduced pro-inflammatory chemokine or cytokine production in the CNS after RABV infection. The findings suggest that Met-CCL5 might be used as a novel therapeutic reagent to prolong the survival time after RABV infection.
Methods
Animals and virus strains
Suckling mice (7-day-old) and adult mice (6-week-old) were purchased from the Institute of Laboratory Animal Medicine at the Chinese Academy of Medical Sciences (CAMS & PUMC, Beijing, China) and housed in the BSL-3 facility of the Veterinary Research Institute at the Academy of Military Medical Sciences. All procedures were conducted in accordance with the guidelines for the Medical Laboratory Animal (1998) from Ministry of Health, China. All animal experiments were carried out as approved by the Institutional Animal Care and Use Committee, Chinese CDC (permission number: 12-0121). Three strains of rabies viruses isolated in China were used in this study, including aG [
32], CTN [
33] and HN10 [
34]. The background for isolation, hereditary traits and genomic information of these strains were well documented. The parental virus of the aG strain was isolated from the brain of a rabid dog in 1931, and it was attenuated to a fixed strain by 30 passages in rabbits, 55 passages in primary hamster kidney cells (PHKC), and several passages in guinea pigs and PHKC [
32]. The aG strain was pathogenic for adult mice by intracerebral (i.c.) inoculation, but non-pathogenic through peripheral infection; it was used as a vaccine in 1981 in China [
32]. The parental virus of the CTN strain was isolated from a rabid patient and attenuated by passaging in mice and human diploid cells. The pathogenic-related amino acid residue in position 333 of glycoprotein was mutated from arginine (R) to glutamine (Q). This CTN strain has been approved as a vaccine strain by the World Health Organization (WHO) since 1983 [
33]. The HN10 strain was a street virus strain isolated from a rabid patient in China in 2006 [
34].
Virus infection and exogenous injection of CCL5 or Met-CCL5
Suckling mice were i.c. infected with aG or CTN. Adult mice were i.c. infected with aG or intramuscularly (i.m.) infected with HN10 (that is, the muscles of the right thigh) (5.6 × 103 FFU, fluorescent focus forming unit; 25 μl in DMEM medium) (n = 10). To assess migration and apoptosis of immune cells infiltrating the CNS, suckling mice were i.c. injected with CCL5 (5 μg/ml, 25 μl), and the same volume of sterile PBS was used for mock controls. For Met-CCL5 treatment, suckling mice were intraperitoneally (i.p.) administrated daily with Met-CCL5 (100 μg/ml, 100 μl) from day 0 postinfection (p.i.) and adult mice were i.p. injected daily with high dose (200 μg/ml, 100 μl), low dose (20 μg/ml, 100 μl) of Met-CCL5, or the control reagent (random sequence of the amino acids from Met-CCL5; 200 μg/ml, 100 μl) beginning on day 2 p.i.. Recombinant carrier-free Met-CCL5 was purchased from R&D (R&D, Minneapolis, MN, USA). Met-CCL5 was derived from E.coli and contained 69aa from Ser24 to Ser91 with an N-terminal Met. The carrier-free control peptide (purity ≥95%) with the same length but random sequence of amino acids from Met-CCL5 was chemically synthesized by Hanhong Chemical Co. Ltd (Hanhong, Shanghai, China).
Preparation of immune cells in spleen or central nervous system
Spleens from RABV-infected mice were homogenized and filtered through a 70-μm nylon cell strainer (Corning, Union City, CA, USA) to prepare single cell suspension. After red blood cells were removed with a lysing solution, cells were prepared for culturing or staining. To isolate immune cells infiltrating the CNS, mouse brains were removed and homogenized using digestion buffer (HBSS containing 0.05% collagenase IV and 10 μg/ml DNase I) and then filtered through a 70-μm nylon cell strainer (Corning, Union City, CA, USA). After digesting at room temperature for 20 minutes, the homogenates were allowed to settle vertically for another 20 minutes [
35]. The clear supernatant was collected and centrifuged, and the single cell solution was prepared for analysis by flow cytometry thereafter.
Flow cytometry
Antibodies (Abs) for flow cytometry were purchased from eBioscience (anti-B220, −CD3, −CD4, −CD11b, −F480, and − CD8), BD Biosciences (anti-CCR5, −Annexin V, and -BrdU), and Cell Signaling Technology (anti-Caspase-3 Asp175). Staining processes were performed according to the manufacturer’s instruction. Briefly, for cell surface straining, cells were blocked with Fc γ MAb (0.5 μg/ml) for 30 min at 4°C. After being washed with PBS, cells were stained with antibodies against B220, CD3, CD4, CD11b, F480, CD8 or CCR5 for 30 min on ice with gentle shaking. For cell apoptosis analysis, fresh cells were stained with anti-Annexin V antibody and propidium iodide (PI), while pre-fixed and permeated cells were stained using anti-caspase-3 antibody. For BrdU incorporation assay, the RABV-infected mice (n ≥3) were i.p. injected with BrdU (10 mg/ml, 100 μl) for 24 hours (hrs); cells from brains were then collected. After being stained with cell surface markers for 30 min on ice, the cells were stained with anti-BrdU antibody according to the manufacturer’s instruction. Samples were processed using FACSCalibur or Accuri C6 (BD Biosciences, San Jose, CA, USA), and data were analyzed with the FlowJo software (Tree Star, Ashland, OR, USA).
Western blot
Western blot was performed using the following primary antibodies: phospho-Akt (Ser 473) and phospho-Fak (Tyr 925) (Cell Signaling Technology, Danvers, MA, USA). Briefly, splenocytes from the RABV-infected and mock-infected mice were harvested, and splenocytes from the mock-infected mice were stimulated with or without 60 ng/ml CCL5 at 37°C for 5 min. After being washed with PBS, the cells were lysed using SDS-loading buffer and boiled for 10 min. Lysates were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred thereafter to a nitrocellulose membrane. After being blocked in TBST with 5% BSA for 1 hour (h), the membrane was incubated with primary antibodies overnight at 4°C, followed by incubation with horseradish peroxidase-labeled secondary antibody. Detection was performed with ECL substrate (Thermo Scientific, West Palm Beach, FL, USA).
Inflammation antibody array assay
Mouse inflammation antibody array G (RayBiotech, Norcross, GA, USA) was used to determine the protein expression levels of various cytokines and chemokines in the brain samples. Briefly, proteins were extracted from brains of the moribund mice infected with aG, CTN or HN10 (n ≥3 in each group). The array was inoculated with 50-μg proteins in 100 μl of the supplied buffer and then treated with biotin-conjugated antibodies for 2 h. After being washed, the array was incubated with Alexa Flour 555-conjugated streptavidin for 2 h at room temperature, and the images were visualized using an Axon GenePix 4300A laser scanner (Molecular Devices, Sunnyvale, CA, USA).
Quantitative real-time PCR
The relative mRNA expression levels of IL-1β, IL-6, IL-12, IL-17, CCL3, CCL5 and virus nucleoprotein were measured by quantitative real-time PCR (qRT-PCR). Briefly, total RNA was isolated from brains of the RABV-infected mice using TRIzol reagent (Invitrogen, Grand Island, CA, USA). cDNAs were synthesized from mRNA by Ready-To-Go You-Prime First-Strand Beads (Amersham Biosciences, Piscataway, NJ, USA) using d(N)6 as primers. qRT-PCR was performed using SYBR Green real-time PCR master mix on the CFX96 system (Bio-Rad, Hercules, CA, USA). To examine virus replication in brains, the mRNA levels of virus nucleoprotein were measured by qRT-PCR using specific primers to detect N protein. The mRNA copy numbers were normalized to the housekeeping gene β-actin.
Preparation of bone marrow-derived macrophages
Bone marrow-derived macrophages (BMM) were generated using L929-cell conditioned medium (LCCM) as a source of granulocyte/macrophage colony-stimulating factor. Briefly, bone marrows were removed from tibias and femur bones of 8-week-old mice. Following red blood cell lysis and washing, cells were plated in DMEM medium (Invitrogen, Grand Island, CA, USA) supplemented with 10% fetal calf serum (CSF), 1% streptomycin/penicillin, and 30% LCCM. On day 7, BMMs were treated for 6 hrs with CCL5 (60 ng/ml), poly (I:C) (polyinosinic-polycytidylic acid), or poly (I:C) and CCL5 with or without Met-CCL5. The relative mRNA levels of cytokines or chemokines were evaluated by qRT-PCR.
Transwell migration assay
Splenocytes were isolated from the RABV-infected mice and their migration ability with or without CCL5 stimulation was subjected to a transwell migration assay. Briefly, 600 μl RPMI1640 medium with or without 100 ng/ml CCL5 was added into a 24-well plate, and then splenocytes (0.5million/100 μl) were placed in the transwell inserts (Corning, Union City, CA, USA). The transparent polyester membranes of the inserts were immersed in the RPMI medium. After being incubated at 37°C for 6 hrs, the cells that had migrated through the polyester membrane into the RPMI medium were collected and counted under an Olympus CK40 light microscope (Olympus, Tokyo, Japan).
Preparation and staining of brain section
RABV-infected mice were euthanized and perfused with PBS followed by 4% paraformaldehyde fixation. The brains were removed and fixed in 4% paraformaldehyde at 4°C for 24 hrs and then immersed in a 10% sucrose solution at 4°C for 48 hrs. Coronal frozen sections of brain tissue (20 μm) were cut on a Leica CM1900 microtome (Leica, Wetzlar, Germany). Sections were stained with cresyl violet to examine the pathological changes of neurons. To detect activated caspase-3, brain sections were incubated with PBS/0.2% Triton for 5 min, and then blocked with PBS containing 5% donkey serum and 0.1% Tween 20 for 2 hrs. They were then stained with rabbit anti-active caspase-3 polyclonal antibody (Promega, Madison, WI, USA) overnight at 4°C. After being washed, brain sections were incubated with Cy3-donkey anti-rabbit IgG antibody (Jackson Laboratories, West Grove, PA, USA) for 2 hrs at room temperature. For terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, the frozen tissue sections were stained according to the manufacture’s protocols for DeadEnd Colorimetric TUNEL System kit (Promega, WI, USA). The average number of apoptotic cells in different regions of hippocampus (CA1, CA3 and DG) was quantified. To detect immune cell infiltration in the brain, T lymphocytes, macrophages, and activated microglia were stained with Alexa Fluor 488-anti-mouse CD3 monoclonal antibody (17A2), anti-mouse integrin αM/CD11b antibody, and Northernlights NL557 fluorescent secondary antibody (R&D, Minneapolis, MN, USA), respectively. The numbers of infiltrated CD3+ T cells in the brain were counted under 10× magnification for each slide. Images were taken from an Olympus FV500 confocal microscope (Olympus, Tokyo, Japan).
Statistical analysis
Statistical significances between different groups were analyzed using a two-tailed Student’s t-test or two-way ANOVA, and statistical significance of survival rates was determined by the log rank test and Kaplan-Meier survival analysis. P <0.05 was considered statistically significant (*P <0.05; **P <0.01; ***P <0.001).
Discussion
RABV is a highly neurotropic virus inducing acute infection in the CNS. Chemokines are now recognized as critical regulators of leukocyte trafficking in the CNS, and numerous studies have revealed that resident cell populations of the CNS are able to synthesize and secrete a variety of chemokines. Neurons, microglia and astrocytes are the primary sources of chemokines following infection with a wide range of neurotropic viruses, including RABV [
42], WNV [
43], and herpes simplex virus 1 (HSV1) [
44].
In vitro studies have highlighted that infection of neurons with RABV results in robust production of chemokines [
42]. In this study, we screened the expression levels of 40 inflammatory molecules in the CNS and demonstrated that CCL5 is the most significantly upregulated chemokine in response to aG and CTN infection in suckling mice, which suggested chemokine CCL5 is one key immune regulator during RABV infection.
Chemokines are critical mediators of neuropathology during viral infections in the CNS, either by attracting pathogenic inflammatory cells or directly mediating neurotoxicity and cell death. Chemokines regulate infiltration of immune cells to the CNS after viral infection by enhancing blood-brain barrier (BBB) permeability [
45]. These immune cells release additional chemokines, such as CCL5, to further recruit macrophages, monocytes and T cells, consequently increasing the severity of demyelination [
23]. A previous study showed that recombinant RABV expressing CCL5 or IP-10 increased pathogenicity with excessive infiltration of inflammatory cells into the CNS [
11]. In this study, excessive infiltration of T cells, macrophages or activated microglia and neuron apoptosis was detected in mice infected with attenuated RABVs, which is in agreement with what others have reported [
5]. Moreover, our
in vivo and
in vitro data suggested that CCL5 played an essential role in enhancing trafficking of immune cells in the CNS during RABV infection.
CCL5 belongs to the CC family of chemokines and plays an important role in recruiting T cells, macrophages and monocytes to the site of inflammation by interacting with the specific G protein-coupled receptors CCR1, CCR3 and CCR5 [
20],[
21]. The binding of CCL5 and its receptors CCR5 activates a series of downstream effectors that facilitate leukocyte trafficking into the CNS [
37]. Although immune cell infiltration and anti-viral activity is requisite for viral clearance, excessive accumulation of leukocytes within the CNS results in neuropathology [
11]. Met-CCL5 is a CCL5 receptor antagonist, which is used extensively to block the pathogenic effects of CCL5 and attenuate CCR5-mediated inflammatory processes during the development of arthritis, colitis, airway inflammation, allograft rejection and chronic liver diseases [
46]-[
50]. Moreover, CCR5 is involved in HIV entry to target cells, so CCR5 antagonists have been successfully tested in phase III studies in patients with HIV infection. In this study, i.p. administration of CCL5 antagonist Met-CCL5 significantly reduced CCL5 and IL-6 in aG-infected suckling mice. Other studies have shown that street RABVs also induce a strong inflammatory response in adult mice or dogs at the late stage of the infection [
16],[
18]. Since upregulated inflammatory cytokines and chemokines were also observed in adult mice infected with street RABV by i.m. route, an investigation as to whether Met-CCL5 was protective in these mice was carried out. Indeed, Met-CCL5 significantly protected adult mice from the infection of the street strain HN10, especially at the early stage of infection. According to previous studies, pretreatment with Met-CCL5 at doses of 0.1 and 1 μg/mouse significantly inhibits cellular recruitment in a murine model of airway inflammation [
49]; furthermore, administration of 10 μg Met-CCL5 for 3 days (30 μg in total) at the peak of liver fibrosis significantly inhibits fibrosis progression and accelerates its regression in chronic liver diseases [
51]. In our study, administration of low dose Met-CCL5 (2 μg daily; 20 μg for aG-infected or 30 μg for HN10-infected adult mice in total) was enough to inhibit RABV infection. This might explain why no significant difference was observed between low (2 μg) and high (20 μg) doses of Met-CCL5 treatment.
However, a limited protective role of Met-CCL5 was observed at the late stage of infection. Considering the complex pathogenic reasons for the lethal RABV infection, no single therapeutic reagent is likely to be effective. Therefore, it is crucial to consider a combination of Met-CCL5 therapy with other treatments to improve protection against RABV. Efforts aimed at developing successful therapeutics to combat RABV infection have been in progress for decades. Despite attempts at drug development and antiviral regimens, there are still no effective treatments for use in the clinic. Several reports have suggested that RABV neutralizing monoclonal antibody cocktails could protect mice from a lethal dose of RABV infection [
52],[
53]. Therefore, if Met-CCL5 treatment could open a therapeutic window at the early stage, it would be fundamental to further investigate whether Met-CCL5 and the RABV neutralizing monoclonal antibody cocktails could be used as a combined therapy to protect the host from RABV infection.
Acknowledgements
This work was supported by grants from the Ministry of Science and Technology of China (2012CB910800), National Natural Science Foundation of China (31070778, 31370859, 31300723), Shanghai Pujiang Program (11PJ1410700), National Department Public Benefit Research Foundation (201103032), Key Technologies Research and Development Program of China (2009ZX10004-705), Instrument Developing Project of the CAS (YZ201339) and funding from the cancer center of Xuhui Central Hospital (CCR2012005). Dr. HW is a scholar of the Hundred Talents Program of the CAS. We thank helpful comments from Dr. Bin Wei (SIBS), thank Drs. Rongliang Hu, Shoufeng Zhang, Jinghui Zhao (Academy of Military Medical Sciences) and Yandao Gong (Tsinghua University) for RABV infection and preparation of the frozen sections.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YH, SJ, WJ, XT, and JX performed experiments and statistical analysis. HW, QT, YH, SJ, and GL participated in the design of the study. HW, QT, YH, and SJ drafted the manuscript. YZ and XX designed primers. All authors read and approved the final manuscript.